We have been working towards creating newroutes to arene compounds by looking at CH activation chemistry. An initial step towards this goalhas now been published as a Journal of the American Chemical Society communication (1). The results have provided us with material for a full paper onthis topic and a related reaction now published in Chemical Communications (2).

In initial work we looked for the cyclometallation-induced formation of M–C bonds to sp²carbons of aryl and vinyl groups. These were successfully achieved for a whole range of substrates.However, the reaction was only successful with iridium, not rhodium. Nevertheless rhodium is tried each time and it is likely that we will get a similar rhodium-induced reaction in the future.

The reaction shown below produces cyclomet-allated species for a series of substrates such as enones and aryl ketones. Attempts to insert CH₂ = CHR species (such as the usually particularlyreactive CH₂ = CHSi(OEt)₃) were unsuccessful, so alkynes were tried. The reaction with alkynes was remarkable, giving a double alkyne to vinylidene rearrangement, followed by an insertion into M–H to give a vinyl, followed by a second alkyne to vinylidene rearrangement and a second insertion. The resulting butadienyl proved to have a ‘π-agostic’ structure with M...C bond lengths much longerthan those of single bonds (2.5 Å was typical). No subsequent treatment, however, has so far given the required coupling.

Deuterium labelling showed that the RC≡CD deuterium 1,2-migrates selectively to provide RDC≡C in the vinylidene. This occurs without oxidative addition because that would lead to a fluxional 7-coordinate M(H)(D) species or more likely to a 6-coordinate M(H–D) species. Both processes are excluded because they would scramble the isotope. The 1,2-migration reaction isstrictly intramolecular within each RC≡CD group,because with RC≡CD and R′C≡CH both presenting the same reaction mixture (crossover reactionconditions), RDC=C and R′HC=C are the only resulting vinylidenes.

In future work, we hope to persuade rhodiumto carry out related coupling reactions, perhaps by moving to the N -heterocyclic carbene ligand system (L′) that has been so successful in alkene metathesis and palladium coupling. Indeed we already have the required [(cod)RhL′CI] species in hand for trials.

Other work (3) includes the synthesis of rhodium and iridium(I) complexes of N -heterocyclic carbenes by transmetallation from the corresponding silver(I) complexes; and the involvement of rhodium in an unusual norbomadiene to nortricyclyl rearrangement.